The subject of this research project is to develop and experimentally validate a novel computational approach to analyze structural ensembles in proteins and peptides that are largely monomeric and unstructured in solution but have the ability to form amyloid fibers. The tools for all-atom computer simulations for such systems are currently underdeveloped which has significant consequences for our understanding of the molecular determinants of amyloidosis. In order to put the development of computational tools in a biological context, we will apply them to a 39-residue peptide from human acidic prostatic phosphatase that is known to form fibrillar structures. These fibrillar structures have been shown to be important for enhancement in viral infectivity of human cells. The computational study will be validated by a battery of experimental methods including NMR, atomic force microscopy, fluorescence and circular dichroism spectroscopies, and hydrogen exchange mass spectroscopy.